Online process particle sizing

Particle size data is important in many industrial applications; either during process development, or else for dry-product analysis.

Analysis of product in dry form is in principle not difficult and many methods for sizing are available, ranging from sieve trays to laser diffraction.

Tracking particles online, while the chemistry is ongoing, is much more difficult and the choice of tools is very limited.

There are several reasons why in-situ analysis is important, though the precise data required is not the same in all cases.

The technically compelling case is that particles removed from the process environment are rarely the same, even before they are exposed to the external measuring technique.

There are also practical limits - taking a representative sample is not always easy, assuming a sample can be taken at all.

Also, samples are needed at times when important events take place and this is certainly not easy to decide in general.

The ability to accurately track particle size and size-distribution during a process (online, for example) can provide an understanding (and hence potentially an improvement) in product quality that is simply not possible by taking samples.

Apart from the obvious advantage of time saving and convenience, the results from offline sample analysis can often be totally different to the original material in the process vessel.

The process conditions - high or low temperature and elevated pressure, for example - can also prohibit sampling.

In fermentation and other biological reactors, particle tracking can be used to monitor increase in biomass, essentially the product of the reaction.

This is traditionally done by taking samples out of the reactor at frequent intervals and then using an off-line counting technique.

This is not only time-consuming but presents an additional problem of losing sterility, leading to total loss of the batch.

Optical density (a form of turbidity) is often used to track bioactivity, but this does not always give sufficiently reliable quantitative information.

In the chemical process industry too, formation of solids is sometimes an integral part of the chemistry, and the ability to monitor this online and use this information for feed-back control can be invaluable.

Many polymerisation processes (such as suspension polymerisation) are examples of this.

Product blending is another application where knowledge of solids and the changes in size and number provides important information regarding the final product properties, and online tracking can substantially reduce product waste and improve quality.

Particle monitoring is also crucial in the control of crystallisation, especially in the pharmaceutical industry, where the FDA has given weight to this through the PAT initiative which encourages more online and ongoing product quality control.

Effectiveness of drugs can be influenced by their crystalline form, and indeed different physical forms of the same chemical are listed in patents.

The ability to monitor particle size is especially useful for troubleshooting in situations where filtration and drying become difficult, or when product quality is a variable.

In many research and industrial applications, it is sufficient to simply track changes in particle size or size distribution, rather than have absolute knowledge.

For example, in the context of crystallisation, the change in particle size profile within different process techniques could be very useful, and sufficient to guide the development in the correct direction.

In tracking biomass during fermentations, it is important to know that there is still activity and if it is increasing or decreasing, without the need for absolute numbers.

At the simplest level, turbidity can provide this type of information and HEL has used turbidity to monitor bioactivity and generate solubility data.

Turbidity rises and falls sharply as a solution is heated and cooled, indicating clearly points of solubility and recrystallisation.

The difference between these two temperatures, called the meta-stable zone width (MSZW), at a range of concentration can easily be generated when integrated into a computer-controlled reactor platform.

When more detailed information about the particles is required, laser reflection techniques are needed; one method is the FBRM from Mettler-Toledo, which is primarily a trending tool that produces statistics in real time, based on particle size and population.

Though it has been correctly described as a 'sophisticated turbidity meter', it has proved very popular due to the fact that it can provide information on qualitative changes (trends) in particle size and population.

Particle sizing by laser reflection works by detection of light bounced off moving solid particles.

Knowing the relative speeds of the particle and light source, and the time to pass from one solid edge to the next, allows the particle size to be calculated.

FBRM uses a rotating laser source to increase the amount of space sampled and provide more reliable figures, but suffers from the fact that it uses signals from a wide range of particles - many of which are not in focus and essentially contribute noise.

As a result, the size information is only indicative rather than quantitative.

A device that substantially reduces the noise, by making use of reflections only from particles in focus, is the Lasentrac, which has the potential to provide much more accurate numbers.

In its simplest implementation, however, described as 'fixed focus' (FF), the final result is only marginally better than FBRM.

This is due to the fact that the sampled particle population at a fixed point can be rather small, especially at dilute concentrations.

The final result is that the simplest version of Lasentrac is generally also good for following trends and changes and not to get absolute numbers, except at higher concentrations.

The lack of moving parts does, however, have the advantage of being much simpler and virtually maintenance-free.

There are many applications where simply trending is not acceptable and greater precision is essential.

Fortunately, a solution is available and involves adding two extra features to the FF version.

The MF (moving focus) version of Lasentrac does this.

The laser source is not static but rotates at high speed, thus a larger number of particles is sampled.

The moving focus enables the depth of focus to be changed, giving 3D movement and the ability to measure a wider particle size range.

For example, particles below 0.2um can be measured and counted, allowing crystallisation kinetics to be tracked from the nucleation stage to final product formation.

The 32-bit processor, used for data handling in all versions of Lasentrac, also helps by enabling the data to be separated into discrete sizes rather than lumped into broad bands where important detail can be lost.

The availability of accurate particle size information online is rather like putting a microscope inside a process vessel - the results have been confirmed by cross-checking with figures from particle image analysers and microscopes.

This is useful in situations where particle size is an important part of the process or product, and if the particle size gives important information about other properties that cannot directly be monitored.

For example, different polymorphs can have different sizes or shapes, and this can be more reliably detected by a correctly configured Lasentrac.